# Laya

> Updated 2026-09-19 · type: tool · category: ai-infrastructure · status: active · rev 1

Laya runs typed classification, scoring and confidence gates locally in one model pass, with separate English, multilingual and task-tuned checkpoints.

- Open source: yes (Apache-2.0)
- Self-hostable: yes
- Pricing model: free
- Best for: Teams routing support, lead, moderation or agent events that need a fast local classifier with explicit probabilities, rather than paying a generative model to produce and then parse text.
- Last verified: 2026-09-19

- **Canonical:** https://gtmstacker.com/registry/tool/laya/
- **Source:** [github · NandhaKishorM/laya](https://github.com/NandhaKishorM/laya)
- **Tags:** ai-infrastructure, classification, model-routing, support-triage, self-hostable
- **Repository:** https://github.com/NandhaKishorM/laya

## Is Laya open source?

Yes, Laya is open source under the Apache-2.0 license.

## How much does Laya cost?

Laya is free to use.

## Can I self-host Laya?

Yes, Laya can be self-hosted (the source is available under the Apache-2.0 license).


---

Apache-2.0 local decision engine for a narrow job generative models are often overused for: take text or JSON, answer a fixed set of typed questions, and return labels, scores and probabilities without generating prose. The Python package exposes choice, ordinal-score and boolean-probability primitives, plus presets for model routing, prompt guardrails, moderation and support-ticket triage. Three open checkpoints cover English, multilingual input and the project's task-tuned workflows.

## Provenance

- Apache-2.0, Python package, open Hugging Face weights, three checkpoints and local inference independently verified from the repo and model card on 2026-09-19 (github.com/NandhaKishorM/laya, ~846 stars, pushed 2026-09-19; convaiinnovations/laya, ~406 Hugging Face likes).
- Surfaced in both lanes: the 2026-09-19 discovery pull carried the launch at 567 HN points, and the viral-X pull carried it as the open alternative to TypeSafe Jev.
- The 33 ms single-question latency, benchmark scores and Jev comparisons are the maker's published T4 results. They were not independently rerun for this registry.

## Why it matters for a GTM stack

Support routing, lead classification, churn flags and prompt screening are decision problems before they are writing problems. Laya gives those steps a local, inspectable probability instead of paying a large model to write a sentence that another rule has to parse. That makes it useful ahead of a CRM, support queue or agent tool call where low confidence should hand the case to a person. The honest read is in the project's own limitations: use the task-tuned checkpoint for its advertised workflows, fit calibration on your data, and do not treat a confident score as proof when the input or label space falls outside what the model handles well.